Cardinal

Cardinal performs statistical analysis of mass spectrometry imaging (MSI) datasets from biological tissue samples, supporting Matrix-Assisted Laser Desorption/Ionization (MALDI) and Desorption Electrospray Ionization-based MSI workflows and analyses of multiple tissue types and complex experimental designs.

Key Features:

• Image Segmentation: Partitions tissue into regions with homogeneous chemical composition, selects an optimal number of segments, identifies informative ions, and characterizes segmentation uncertainty.
• Image Classification: Assigns spatial locations on tissue samples to predefined classes, selects the most informative ions for classification, and estimates classification error using cross-validation.
• MSI Modality Support: Supports Matrix-Assisted Laser Desorption/Ionization (MALDI) and Desorption Electrospray Ionization-based MSI workflows.
• Experimental Design Support: Handles experiments involving multiple tissue types and complex experimental designs.
• Statistical Framework: Implements mixture modeling and regularization to model complex data structures and improve model accuracy.

Scientific Applications:

• Oncology: Analysis of tissue-derived MSI datasets relevant to oncology research.
• Pathology: Spatial chemical analysis of tissue samples for pathology investigations.
• Pharmacology: Examination of tissue molecular distributions in pharmacology studies.

Methodology:

Uses mixture modeling and regularization; selects informative ions and optimal segment number; characterizes segmentation uncertainty; and estimates classification error via cross-validation.